Checking cell size in yeast

Review

TRENDS in Genetics Vol.18 No.9 September 2002

47 Zhaxybayeva, O. and Gogarten, J.P. (2002) Bootstrap, Bayesian probability and maximum likelihood mapping: exploring new tools for comparative genome analyses. BMC Genomics 3, 4 48 Page, R.D. (2000) Extracting species trees from complex gene trees: reconciled trees and vertebrate phylogeny. Mol. Phylogenet. Evol. 14, 89–106

49 Bapteste, E. et al. (2002) The analysis of 100 genes supports the grouping of three highly divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba. Proc. Natl. Acad. Sci. U. S. A. 99, 1414–1419 50 Woese, C.R. (2000) Interpreting the universal phylogenetic tree. Proc. Natl. Acad. Sci. U. S. A. 97, 8392–8396

479

51 Burggraf, S. et al. (1991) Methanopyrus kandleri: an archaeal methanogen unrelated to all other known methanogens. Syst. Appl. Microbiol. 14, 346–351 52 Slesarev, A.I. et al. (2002) The complete genome of hyperthermophile Methanopyrus kandleri AV19 and monophyly of archaeal methanogens. Proc. Natl. Acad. Sci. U. S. A. 99, 4644–4649

Checking cell size in yeast Ivan Rupeš To remain viable, cells have to coordinate cell growth with cell division. In yeast, this occurs at two control points: the boundaries between G1 and S phases, also known as Start, and between G2 and M phases. Theoretically, coordination can be achieved by independent regulation of growth and division, or by participation of surveillance mechanisms in which cell size feeds back into cell-cycle control. This article discusses recent advances in the identification of sizing mechanisms in budding and in fission yeast, and how these mechanisms integrate with environmental stimuli. A comparison of the G1–S and G2–M size-control modules in the two species reveals a degree of conservation higher than previously thought. This reinforces the notion that internal sizing could be a conserved feature of cell-cycle control throughout eukaryotes. Published online: 25 July 2002

Ivan Rupeš Dept of Biology, Queen’s University, Kingston, Ontario, Canada K7L 3N6. e-mail: rupesi@ biology.queensu.ca

How does a cell control its size? Two schools of thought exist: either a cell divides after it reaches a certain critical size, or cell growth and proliferation are regulated independently, with cell size emerging from a simple correlation of the two [1–5]. The most important difference between the two hypotheses is that cell size feeds back into the cell-cycle regulatory system in the former but not in the latter (Fig. 1a). Thus, the validity of the critical size theory depends on the existence of a ‘sizer’ – a molecule or set of molecules whose activity correlates with cell size. Nevertheless, the sizer is only one component that determines when cell division occurs; in addition, the extracellular environment influences the timing of the response to the changing activity of the sizer. Two species of yeast, the budding yeast Saccharomyces cerevisiae and the fission yeast Schizosaccharomyces pombe, provide genetic models in which to study cell-cycle control. The evolutionary divergence of these yeast is about the same as that between each of them and human [6]. Both yeast share cell-cycle characteristics with higher eukaryotes, such as G1, S, G2 and M phases, cyclindependent kinases (CDKs) and checkpoint controls [7–9]. Owing to the uncertain supply of nutrients in the wild, the yeast cell-division rate must be coordinated with widely variable rates of cell growth, otherwise cells would get progressively smaller or larger. Although cell size is reduced in less favorable http://tig.trends.com

growth conditions, the range of growth rates exceeds the corresponding range of cell sizes for both yeast. Thus, a relationship between growth and division is a fact of life for yeast cells. The early studies have presented a compelling case for the existence of a critical size that is a prerequisite for progression through the cell cycle. In budding yeast, cell division is asymmetrical and produces cells of unequal size. To compensate for this asymmetry, which becomes more pronounced with increasing nutrient limitation, new daughter cells grow more before division than the mother cells. This additional growth occurs almost entirely in G1, before the reference point known as Start. Once Start is passed, the rest of the cell cycle is relatively constant in length [10] (Fig. 1b). In fission yeast, G2–M is the primary cell-size control point [9]. The relative length of G2 varies greatly with growth conditions (Fig. 1b). Nitrogen limitation reduces cell size at division, and sudden shifts between different sources of nitrogen generate rapid acceleration or delay of mitosis in cells that are above or below the new cell-size threshold (i.e. the minimum size required for initiation of mitosis) [9]. Even during balanced growth, individual fission yeast cells can compensate for random fluctuations in their size at birth by adjusting their time spent in G2 [11]. Do fission yeast have a size control in G1, and do budding yeast have a control in G2? Start is defined in fission yeast as in budding yeast, although in favorable growth conditions it occurs almost immediately after exit from mitosis. A normally cryptic size-control point at Start is uncovered in mutants in which the G2–M size-control has collapsed and cells enter mitosis prematurely. These cells have an extended G1, suggesting that they initiate S phase only after reaching a certain minimum size [12] (Fig. 1b). Although the two yeast have traditionally been thought of as using different size-control strategies, the aim of this article is to show that they use similar mechanisms. What is different is their emphasis on the size-control points. Because these control mechanisms have been conserved over such long evolutionary distances,

0168-9525/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0168-9525(02)02745-2

480

Review

TRENDS in Genetics Vol.18 No.9 September 2002

(a)

Glucose

(Sizer)

External conditions (nutrition)

4 Cell-cycle progression 3 Growth rate

1 2

Grr1

Cln3 Cdc28

Swi6 Swi4

Budding yeast Wild type, rich medium G1

Cln1

Daughter cell

S

G2 M Mother cell

G2

Cln2 Cdc28

Cdc28

Cdc28 (Inactive) P

M Daughter cell

G1

Clb5 Cdc28

S

G2 Mother cell

M

P

Sic1

Sic1

Wild type, poor medium

Budding, spindlepole duplication

Sic1 Clb6

Sic1 (Inactive) Clb5 Cdc28

Cln3 over production, rich medium S

Size

cAMP

Cell size

(b)

G1

Growth

Clb6 Cdc28

Swi6 Mbp1

Fission yeast G1

Wild type, rich medium G1

S

G2

wee1, rich medium G1 S G2 Wild type, poor medium G1 S

S

M

M

Size

G2

M TRENDS in Genetics

G1 cyclin

CDK1

B-type cyclin

Transcription factor

CDK inhibitor TRENDS in Genetics

Fig. 1. External conditions and internal sizing mechanisms co-determine cell size at division. (a) Integration map of the cell-size control modules. Cell size is a result of correlation between the rates of cell-cycle progression (proliferation) and growth, but four different general processes modulate the two rates. (1) Cell-cycle machinery monitors external conditions or activation of metabolic pathways. (2) Another class of mechanisms sets the growth rate depending on external conditions. (3) Growth rate can directly modulate cell-cycle progression. (4) Cell size feeds back into cell-cycle control through a specific size-surveillance mechanism (sizer). (b) Durations of individual cell-cycle phases in budding and fission yeast. Budding yeast: in wild-type conditions, smaller daughter cells undergo substantial growth before G1–S (Start), whereas larger mother cells rapidly initiate S phase. G1 is drastically reduced when Cln3 is overexpressed, resembling the wild-type cycle of fission yeast, whereas the S phase remains virtually unchanged. G1 becomes relatively extended in poor growth conditions, and S, G2 and M are not affected. Fission yeast: cells initiate S phase rapidly after the completion of mitosis. Cells defective in wee1 enter mitosis prematurely. Their G1, however, becomes extended but their S phase remains unchanged, resembling the proportions in wild-type budding yeast. In poor conditions, G1 also becomes relatively extended in fission yeast.

it might be worth considering whether analogous systems operate in other eukaryotes.

Fig. 2. The G1–S cell-size control module in budding yeast. A simplified minimum core consists of three G1 cyclins, Cln1–3, two B-type cyclins, Clb5 and Clb6, and a cyclin-dependent kinase (CDK) inhibitor Sic1. All the cyclins function in association with a CDK1 homolog, Cdc28. Cln3 is placed at the top of the cascade. In a dose-dependent manner, Cln3 induces expression of CLN1 and CLN2 through activation of a transcription activator complex that comprises two subunits: Swi6 and Swi4. This results in a sharp spike of Cln1– and Cln2–Cdc28 activity that initiates post-Start events, including budding and pre-mitotic spindle-pole duplication. The instability of Cln1 and Cln2 is increased through the action of Grr1 (a protein that promotes degradation of Cln1 and Cln2) in response to glucose. In addition, cAMP represses CLN1 transcription. Cln1– and Cln2–Cdc28 phosphorylate and destabilize Sic1, a potent inhibitor of Clb5– and Clb6–Cdc28. Transcription of CLB5 and CLB6 is also cell-cycle regulated through a complex consisting of Swi6 and a Swi4 homolog, Mbp1. This regulation is not rate-limiting for Start. Activation of Clb5– and Clb6–Cdc28 leads to initiation of DNA replication. Mitotic cyclins are also present in G1 but they are rapidly degraded through the action of Cdh1, a homolog of Ste9 from fission yeast. In contrast to Cdc13 in fission yeast, the budding yeast mitotic cyclins are unable to initiate DNA replication efficiently, and are therefore omitted from this diagram. For simplicity, regulation of the individual components at the transcriptional, post-transcriptional and degradation levels is not distinguished in this diagram. Coloring reflects structural relationships as indicated.

Size control at Start: budding yeast

How does budding yeast measure its size, and how does this control Start? The G1–S size-control module is presented in Fig. 2. Its core consists of three G1 cyclins (Cln1–3) and two B-type cyclins (Clb5 and Clb6). These cyclins bind to a CDK1 homolog, Cdc28 (Table 1), and target its activity towards http://tig.trends.com

specific sets of substrates. The level of Cln3 is important for timely expression of the CLN1 and CLN2 genes. The resulting increase in the activity of Cln1– and Cln2–Cdc28 complexes is responsible for initiation of three major post-Start events: budding, spindle-pole duplication and, indirectly through

Review

TRENDS in Genetics Vol.18 No.9 September 2002

release of the inhibition of Clb5– and Clb6–Cdc28 by the CDK inhibitor Sic1, DNA replication [8,13]. Cln3 functions to coordinate Start with cell growth and the supply of nutrients. The level of Cln3 varies dramatically with the available carbon [14] or nitrogen [15] source. The glucose-response elements in the CLN3 promoter are a target for Azt1, a transcription factor that stimulates CLN3 transcription in the presence of glucose [16,17]. Cln3, therefore, is associated with process 1 in Fig. 1a. CLN3 mRNA contains a long 5′-untranslated region (UTR) that harbors an upstream open reading frame (ORF) whose function is to slow down initiation of CLN3 translation when fewer ribosomes are available in the cytoplasm. Because the concentration of ribosomes correlates with growth rate, this mechanism provides a link between growth rate and the rate of Cln3 synthesis [18]. Thus, Cln3 is also associated with process 3 in Fig. 1a. The translation efficiency of CLN3 is further controlled by the activity of the eukaryotic initiation factor (eIF) 4F mRNA-cap-binding complex. This complex appears to be a target for the target of rapamycin (TOR) phosphatidylinositol kinase pathway [19,20] and perhaps also for the cAMP-dependent protein kinase pathway [14]. Both these pathways are involved in nutritional sensing and affect several metabolic and growth-related functions, thus coordinating processes 1 and 2 (Fig. 1a). Finally, the Whi3 protein restricts Cln3 synthesis by localizing CLN3 mRNA to specific cytoplasmic loci with a possible function in nutritional sensing (Ref. [21] and B. Futcher, pers. commun.). Is Cln3 a sizer? Cln3 is a low-abundance, constitutively unstable protein [22,23] and, therefore, its level depends primarily on its rate of synthesis. According to one model [13], the rate of Cln3 synthesis per cell increases with the volume of cytoplasm. Because Cln3 localizes to the nucleus [24,25], assuming that nuclear volume depends on the amount of DNA present, the effective concentration of Cln3 in G1 rises with increasing cell size. When a threshold level is reached, which takes longer in smaller daughter cells than in mother cells, the cascade leading to Start is initiated [13]. The short answer to the above question is, therefore, yes, Cln3 is most likely a sizer. There is some uncertainty, however, about how exactly Cln3 might do the sizing. For one, it is not clear how a steady increase in Cln3 levels in G1 could trigger the sharp rise in the levels of Cln1 and Cln2 with little apparent contribution from a positive amplification loop [22–27]. One could imagine that the transcription factor complex Swi4–Swi6 (Fig. 2) operates as an ultra-sensitive switch that produces a steep activation curve once the Cln3–Cdc28 level crosses the threshold value [28]. If so, one would expect excessive variation in cell size caused by minor random fluctuations of the Cln3 levels in individual cells; this is not the case. The level of CLN3 transcription oscillates modestly during the cell cycle, peaking in late M or G1 [29,30]. Recently, new evidence suggested that, in cooperation http://tig.trends.com

481

with SWI4, this modest periodicity might, after all, contribute to more robust timing of Start. The upstream-activating sequences of CLN3 and SWI4 contain early cell-cycle box (ECB) elements that control the periodicity of the respective transcripts [29,30]. Cells lacking functional ECBs in both CLN3 and SWI4 also lack the sharp peaks of CLN2 transcription. They also exhibit additive delay and increased size heterogeneity at Start [30], an indication that progression through Start has turned into a stochastic process. Importantly, the peaks of CLN3 and SWI4 transcription seem to occur later in daughter cells than in mother cells, suggesting that their timing might be linked to cell size [30]. However, it is not yet clear whether the timing of the peaks is crucial or whether the cln3ecb swi4ecb double-mutant cells delay Start simply because they fail to produce sufficient levels of Cln3 and Swi4. If cells can sense their size, at least in part, through Cln3, what sets the critical threshold of its activity? Clearly, the threshold varies in different growth conditions. Part of the problem is that the exact link between Cln3 and Swi6–Swi4 is not known [31,32]. What is known, however, is that both synthesis and degradation of Cln1 and Cln2 are modulated by several other pathways. Somewhat counter-intuitively, the burst of CLN1 and CLN2 transcription is delayed in rich media, resulting in a larger size at Start. This response is, at least in part, mediated by cAMP [33,34]. Addition of glucose or cAMP represses transcription of CLN1, and this is sufficient to delay Start [35]. In addition, Grr1, a protein that promotes degradation of Cln1 and Cln2, is activated in response to glucose [36,37]. This further reduces the levels of Cln1 and Cln2 in cells. This type of regulation allows the response of Cln1 and Cln2 to be sensitized or desensitized to Cln3 activity. Cln1 and Cln2 thus make a specific contribution to the integration of environmental signals within the G1–S module (Fig. 1a, process 1) and, therefore, to the setting of a critical size. Size control at Start: fission yeast

In fission yeast, Start is initiated by a CDK1 homolog, Cdc2, in association with a G1 cyclin, Puc1, and three B-type cyclins, Cig1, Cig2 and Cdc13. Two inhibitory mechanisms prevent premature initiation of S phase: one involves a CDK inhibitor, Rum1, and the other involves a component of the ubiquitin proteolysis machinery, Ste9. The wiring diagram reveals striking similarity between fission yeast and budding yeast (Fig. 3, Table 1) [9]. As already mentioned, the fission yeast G1–S size control is cryptic when cells grow in favorable conditions. Under limiting conditions, however, G1 is extended to a varying degree, with no simple correlation to G2–M regulation. This suggests that, as in budding yeast, there might be a minimum size requirement for progression through Start that varies with external conditions [38]. How cell size impinges on the G1–S module in fission yeast is still

482

Review

TRENDS in Genetics Vol.18 No.9 September 2002

Nitrogen

Growth

Size ? P Puc1 Cdc2

?

Ste9

?

? Cig1

(Inactive) Rum1

(Inactive) Rum1 Cdc13

Cig2 Cdc2

Cig1

Cdc2 P

P Rum1

Cdc10 Res1

Ste9

Rum1

Cig2 Cdc2

Cdc13

Cdc13

Cdc2

Cdc10 Res2

G1

S

Size G1 cyclin

CDK1

B-type cyclin

Transcription factor

CDK inhibitor TRENDS in Genetics

Fig. 3. The G1–S cell-size control module in fission yeast. In the simplified version, there is one G1 cyclin homolog, Puc1, and three B-type cyclins, Cig1, Cig2 and Cdc13, active in G1–S in association with a cyclin-dependent kinase (CDK)-1 homolog, Cdc2. Two inhibitory activities are crucial for preventing premature initiation of S phase: a CDK inhibitor, Rum1 (a homolog of Sic1 in budding yeast), and Ste9, a protein involved in targeting its substrates for degradation. Puc1 functions upstream of the B-cyclin–Cdc2 complexes. In G1, Rum1 inhibits Cdc13–Cdc2 and, to a lesser extent, Cig2–Cdc2. In addition, the levels of Cdc13 and Cig1 are kept low by the action of Ste9. At Start, Puc1 and Cig1 phosphorylate and destabilize Rum1, possibly with increasing contribution from Cig1 as the level of Ste9 decreases. Cdc13 phosphorylates Ste9, resulting in Ste9 degradation. This creates a negative loop that contributes to gradual accumulation of Cdc13 during S phase. Transcription of cig2 is induced by a heteromeric complex that contains a Swi6 homolog, Cdc10, and two Swi4/Mbp1 homologs, Res1 and Res2. This transcription is not rate-limiting for Start but is present in this diagram as a feature conserved between the two yeast. Coloring reflects structural relationships as indicated.

largely a mystery. Puc1 alone is insufficient to initiate S phase, indicating that it functions upstream of the B-cyclins. Puc1–Cdc2 phosphorylates and destabilizes Rum1, an inhibitor of Cdc13– and Cig2–Cdc2, paralleling the action of Cln1– and Cln2–Cdc28 on Sic1 in budding yeast [39]. Similar to CLN3 mRNA, puc1 mRNA contains a long 5′-UTR and upstream ORFs, suggesting tight regulation at the translational level [40]. These properties and the low abundance and high instability of puc1 mRNA portray Puc1 as a counterpart of Cln3 in fission yeast. However, in contrast to Cln3, Puc1 seems completely dispensable for the timing of Start in rich medium. By contrast, in poor nitrogen sources, Puc1 becomes rate-limiting for Start [39]. Puc1 might thus function as a G1–S sizer in fission yeast, but this function is http://tig.trends.com

probably shared with additional components. An equivalent of Cln1 and Cln2 is conspicuously missing downstream of Puc1, but so is the extra requirement to coordinate budding with the initiation of S phase for which Cln1 and Cln2 are responsible. Puc1 adopts the role of Cln1 and Cln2 in relieving the Rum1 (Sic1)mediated CDK inhibition, but there appears to be no intermediate Cdc2-dependent transcription step in this cascade [41], and Cig2 activity occurs in a sharp spike that is reminiscent of those of Cln1 and Cln2 [42,43]. The two Clns participate in the modulation of critical size in response to the environment. One might, therefore, expect a similar function for Cig2. Indeed, cig2 mRNA contains long UTRs and translational regulation reduces the levels of Cig2 in response to nitrogen limitation (Fig. 1a, process 1), delaying progression through Start [44]. Although puc1 mRNA is induced by nitrogen starvation [40], the protein levels do not change dramatically (S. Moreno, pers. commun.). Even so, with the relative drop in Cig2 levels, this could contribute to the increased impact of puc1 deletion on cell size at Start that is observed under nutrient limitation [39]. Size control at G2–M: fission yeast

The wee1 gene was isolated as the prototype cell-size mutant that enters mitosis precociously and loses the wild-type ability to modify cell size at mitosis in response to the availability of nitrogen [9]. As cells progress through S and G2, Cdc13 slowly accumulates, but the associated Cdc2 activity remains inhibited by Wee1-dependent tyrosine phosphorylation (Fig. 4a). A Wee1 inhibitory kinase, Cdr1, is probably a component of the nutrition-sensing module (Fig. 1a, process 1) as cdr1-deficient mutants are unable to adjust cell size over a range of nitrogen concentrations [9,45]. This property is shared with a related kinase, Cdr2, which also phosphorylates Wee1 in vitro [45–47]. The exact relationship between Cdr1 and Cdr2 and their possible upstream regulators is not known. Cdr1 localizes predominantly to the cytoplasm, whereas Wee1 is present mainly in the nucleus. Controlled accessibility of the two partners, therefore, is one possible mechanism by which Cdr1 could exert control over the setting of cell size [48]. Independent of Cdr1 and Cdr2, Wee1 is upregulated by inhibition of protein synthesis [49]. To initiate mitosis, cells must reverse Cdc2 tyrosine phosphorylation. The essential phosphatase for this is Cdc25 [9] (Fig. 4a). Switch-like activation of Cdc13–Cdc2 is ensured by a positive feedback loop between Cdc2 and Cdc25 [50] and probably also by a negative loop between Cdc2 and Wee1, the latter of which is related to the Cdc2–Wee1 loop identified in other eukaryotes [51]. The level of Cdc25, as well as that of Cdc13, is correlated with general synthetic activity through translational regulation involving the 5′-UTRs of the cdc25 and cdc13 mRNAs, providing a potential link between growth rate and timing of mitosis [52] (Fig. 1a, process 3).

Review

TRENDS in Genetics Vol.18 No.9 September 2002

483

a,b

Table 1. Conserved components of the G1–S and G2–M size-control modules Budding yeast

Fission yeast

Protein function

Cln1, Cln2, Cln3 Clb1, Clb2, Clb5, Clb6 Cdc28 Sic1 Cdh1 Swi6 Swi4, Mbp1 Swe1 Mih1 Hsl1, Kcc4, Gin4

Puc1 Cig1, Cig2, Cdc13 Cdc2 Rum1 Ste9 Cdc10 Res1, Res2 Wee1 Cdc25 Cdr1, Cdr2

G1 cyclin B-type cyclin CDK1 CDK inhibitor APC activator Transcription factor Transcription factor Protein tyrosine kinase Protein tyrosine phosphatase Protein kinase

(a)

Nitrogen

Size

(Inactive) Cdr1

Cdr2

Cdc13

Y- P

Cdc2 Cdc25

Wee1 Cdc13

?

Cdc2 a

Abbreviations: APC, anaphase-promoting complex (part of the ubiquitin proteolytic machinery); CDK, cyclin-dependent kinase. b Cdc (cell-division cycle) genes were first identified as mutants that cause cell-cycle arrest in budding and fission yeast. The conserved components display a significant degree of structural homology and, mostly, they also have conserved interacting partners, upstream regulators or targets, suggesting that they constitute conserved regulatory modules. The context in which these modules operate, however, can differ in the two yeast, sometimes leading to different phenotypic manifestations in the respective mutants.

G2

M

Size

So, which components qualify for the size-dependent process (Fig. 1a, process 4)? One possibility is that gradual accumulation of Cdc13, which destabilizes the ‘off’ state of the switch, represents a sizing mechanism [53,54]. However, there seems to be little experimental evidence supporting this because accumulation of Cdc13 in the nucleus appears to level off long before cells enter mitosis [55]. The same seems to be true for the pre-mitotic nuclear accumulation of Cdc25 [56]. Wee1 is also an unlikely candidate as Cdc2-tyrosine phosphorylation persists even after Wee1 and a related kinase, Mik1, have been inactivated in a conditional mutant [57,58]. That leaves us with a direct activation of the Cdc25–Cdc2 loop. Indeed, the sizing in G2–M has recently been shown to be mediated by Cdc25. Cells that have stopped growing because of interference with the actin cytoskeleton fail to initiate mitosis if their size is below the threshold, but do progress into mitosis unperturbed if they are oversized [58]. The inhibition is lost in cells in which Cdc25 has been replaced with a constitutive Cdc2-tyrosine phosphatase activity. Conversely, cells in which the critical size has been reduced by a nutritional down-shift accelerate their entry into mitosis even if their further growth has been interrupted by actin depolymerization. These cells suddenly became oversized relative to the new size threshold [58]. The size-related upstream regulators of Cdc25 (or Cdc13–Cdc2, as the presence of the positive loop makes it difficult to distinguish between direct activation of Cdc13–Cdc2 and Cdc25) remain to be identified. However, these results demonstrate that cell size impinges on cell-cycle regulation as a cell-cycle checkpoint. The identification of the checkpoint has an important consequence: it justifies the integration model of cell-size control by closing the loop between cell size and cell-cycle progression (Fig. 1a, process 4). It shows that the activity of the sizer does not simply coincide with cell growth but is http://tig.trends.com

(b) Bud

Nutrition

Size ?

?

Hsl1

(Inactive)

Hsl7

Clb2

Y- P Cdc28 Mih1

Swe1 Clb2

?

Cdc28

G2

M

Size TRENDS in Genetics

Fig. 4. The G2–M size-control module. (a) Fission yeast: the essential B-cyclin–CDK (cyclin-dependent kinase) complex, Cdc13–Cdc2, is inhibited by phosphorylation of Tyr15. The phosphorylation is mediated by Wee1 and reversed by Cdc25. Two feedback loops are engaged, resulting in a switch-like activation of Cdc2 kinase. Cdr1 and Cdr2 are upstream regulators of Wee1 that are involved in nutritional signaling. (b) Budding yeast: the Clb1– and Clb2–Cdc28 complexes share the mitosis-promoting CDK activity, Clb2 being the dominant of the pair (the interactions of Clb1 parallel those of Clb2 and, therefore, are omitted from this diagram). Swe1 and Mih1, the respective homologs of Wee1 and Cdc25, modulate phosphorylation of Cdc28 Tyr19, the equivalent of Cdc2 Tyr15. Hsl1 is a homolog of Cdr1 and Cdr2, and Hsl7 is its interacting partner.

directly determined by cell size. The checkpoint function cannot be executed without previous Cdc2tyrosine phosphorylation. Consistently, wee1-deficient mutants are unable to control their size in G2–M. Furthermore, simultaneous collapse of the G1–S module in the wee1 background, owing to deletion of rum1 or ste9, results in a loss of viability [59–61].

484

Review

TRENDS in Genetics Vol.18 No.9 September 2002

Size control in G2–M: budding yeast

Acknowledgements I thank E. Boye, L. Breeden, B. Futcher, D. Kellogg, D. Lew, S. Moreno and A. Sveiczer for fruitful discussions and for sharing their data before publication. I also thank P. Young, J. Karagiannis and D. Kellogg for critical reading of the manuscript.

In G2, Cdc28 is phosphorylated by the Wee1 homolog Swe1, and dephosphorylated by the Cdc25 homolog Mih1 [8] (Fig. 4b, Table 1). Unlike their counterparts in fission yeast, however, neither of these components are essential in budding yeast. It has been proposed that instead of size control, Clb2–Cdc28 tyrosine phosphorylation evolved a specialized function in execution of a morphogenesis checkpoint. The checkpoint is activated when cells fail to form a bud or when the integrity of the actin cytoskeleton is perturbed [62]. Part of this signal is believed to be mediated by inhibition of Hsl1, one of the three Cdr1 and Cdr2 homologs in budding yeast (Table 1), and its interacting partner, Hsl7. Inhibition of Hsl1 and Hsl7 is thought to result in stabilization of Swe1 and inhibition of mitosis [63]. However, a more sensitive assay revealed that Swe1 is, in fact, normally present in cells until the end of mitosis, just as in other eukaryotes [64]. This revives the question as to whether other functions of Wee1 and Cdc25 homologs, such as size control, might be conserved in budding yeast. Overexpression of SWE1 results in pre-mitotic arrest, and deletion of MIH1 increases cell size [65,66]. Conversely, deletion of SWE1 causes subtle but consistent reduction of cell size (D. Kellogg, pers. commun.). This alone does not tell us whether Swe1 and Mih1 actively participate in size control, but indirect evidence strongly suggests that G2–M size-control is, indeed, present in budding yeast, although it normally remains cryptic. Reversing the situation seen in fission yeast, budding yeast cells that have their G1–S size-control compromised, such as Cln3 overproducers, are small, but their generation time (i.e. time between two subsequent cell divisions) remains largely unaffected, suggesting that the second control-point becomes active and prevents catastrophic regression of cell size [67,68] (Fig. 1b). At the molecular level, Clb2–Cdc28 and Swe1 are engaged in a strong negative feedback loop, suggesting the presence of the same switch-like mechanism of Cdc28 activation that is conserved throughout eukaryotic cells (D. Kellogg, pers. commun.). In light of these results, the concept of a morphogenesis checkpoint in budding yeast requires reexamination to separate the specific response to failed bud formation from a simple manifestation of cell-size control. The effect of cell size has not been directly addressed in this context. It seems, however, that the inability of cells to reach a sufficient size for G2–M might explain the pre-mitotic arrest that is caused by perturbation of actin integrity, as occurs in fission

References 1 Neufeld, T.P. and Edgar, B.A. (1998) Connections between growth and the cell cycle. Curr. Opin. Cell Biol. 10, 784–790 2 Polymenis, M. and Schmidt, E.V. (1999) Coordination of cell growth with cell division. Curr. Opin. Genet. Dev. 9, 76–80 3 Conlon, I. and Raff, M. (1999) Size control in animal development. Cell 96, 235–244 4 Stocker, H. and Hafen, E. (2000) Genetic control of cell size. Curr. Opin. Genet. Dev. 10, 529–535

http://tig.trends.com

yeast [58]. It would be premature to speculate about the identity of the G2–M sizer in budding yeast, but it would be intriguing to test whether some of the Cdr1 and Cdr2 homologs (Table 1) can mediate nutritional signals. If the G2–M module is conserved, what makes Mih1 non-essential? One explanation is that, in the absence of Mih1, Cdc28 inhibition is eventually overturned by continued accumulation of Clb2, activating more and more free Cdc28 molecules [23] until the switch is flipped. Concluding remarks

What is ‘cell size’? Entirely in accord with Pringle and Hartwell [10], I use this deliberately vague expression instead of more specific terms simply because we are still far from fully understanding what represents cell size within cells. Production of Cln3 is clearly dependent on the total number of ribosomes in the cytoplasm, which increases as cells grow, but other possible sizers do not offer such clear-cut solutions. Protein mass, the number of ribosomes and cell volume are obvious candidates, and one or another blend that includes these, and perhaps other structurally based determinants, might constitute an integrated measure of size. So, why did cells bother to evolve such a complicated system of size monitoring and feedback controls if, intuitively, a simple correlation of growth and proliferation could suffice? One possible answer is that cells need it to boost robustness of the cell-cycle regulatory network. There is always variation in the size at which cell-cycle transitions occur. If unchecked, the variation would increase from generation to generation until the cells at the tails of the distribution would enter territory that is incompatible with survival. The cell-size homeostasis phenomenon that prevents this from happening has been demonstrated elegantly in fission yeast [11]. Stochastic processes, to which low-abundance constituents are especially vulnerable, are inherent to biological systems. Even in eukaryotic cells, gene expression can occur in erratic bursts rather than smoothly [69]. Eroding components of the regulatory network causes increased heterogeneity in cell size and unpredictable behavior [30,70]. This leads to the argument that cell-size feedback mechanisms or cell-size checkpoints are a conserved solution to the noise problem. They might underlie more complex mechanisms that have evolved from preexising building blocks such as the ones that constitute processes 1–3 in Fig. 1a to ensure proper development and differentiation in multicellular organisms.

5 Coelho, C.M. and Leevers, S.J. (2000) Do growth and cell division rates determine cell size in multicellular organisms? J. Cell Sci. 113, 2927–2934 6 Sipiczki, M. (2000) Where does fission yeast sit on the tree of life? Genome Biol. 1, 1011.1–1011.4 7 Nurse, P. (2000) A long twentieth century of the cell cycle and beyond. Cell 100, 71–78 8 Lew, D. et al. (1997) Cell cycle control in Saccharomyces cerevisiae. In The Molecular and Cellular Biology of the Yeast Saccharomyces. Vol.III.

Cell Cycle and Cell Biology (Pringle, J.R. et al., eds), pp. 607–695, Cold Spring Harbor Laboratory Press 9 MacNeill, S.A. and Nurse, P. (1997) Cell cycle control in fission yeast. In The Molecular and Cellular Biology of the Yeast Saccharomyces: Vol.III. Cell Cycle and Cell Biology (Pringle, J.R. et al., eds), pp. 697–763, Cold Spring Harbor Laboratory Press 10 Pringle, J.R. and Hartwell, L.H. (1981) The Saccharomyces cerevisiae cell cycle. In The Molecular Biology of the Yeast Saccharomyces: Life Cycle and Inheritance

Review

11 12

13 14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

(Strathern, J.N. et al., eds), pp. 97–142, Cold Spring Harbor Laboratory Press Sveiczer, A. et al. (1996) The size control of fission yeast revisited. J. Cell Sci. 109, 2947–2957 Fantes, P. and Nurse, P. (1978) Control of the timing of cell division in fission yeast. Cell size mutants reveal a second control pathway. Exp. Cell Res. 115, 317–329 Futcher, B. (1996) Cyclins and the wiring of the yeast cell cycle. Yeast 12, 1635–1646 Hall, D.D. et al. (1998) Regulation of the Cln3–Cdc28 kinase by cAMP in Saccharomyces cerevisiae. EMBO J. 17, 4370–4378 Gallego, C. et al. (1997) The Cln3 cyclin is down-regulated by translational repression and degradation during the G1 arrest caused by nitrogen deprivation in budding yeast. EMBO J. 16, 7196–7206 Parviz, F. and Heideman, W. (1998) Transcriptional regulation of CLN3 expression by glucose in Saccharomyces cerevisiae. J. Bacteriol. 180, 4508–4515 Newcomb, L.L. et al. (2002) AZF1 is a glucosedependent positive regulator of CLN3 transcription in Saccharomyces cerevisiae. Mol. Cell. Biol. 22, 1607–1614 Polymenis, M. and Schmidt, E.V. (1997) Coupling of cell division to cell growth by translational control of the G1 cyclin CLN3 in yeast. Genes Dev. 11, 2522–2531 Berset, C. et al. (1998) The TOR (target of rapamycin) signal transduction pathway regulates the stability of translation initiation factor eIF4G in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U. S. A. 95, 4264–4269 Danaie, P. et al. (1999) CLN3 expression is sufficient to restore G1-to-S-phase progression in Saccharomyces cerevisiae mutants defective in translation initiation factor eIF4E. Biochem. J. 340, 135–141 Garí, E. et al. (2001) Whi3 binds the mRNA of the G1 cyclin CLN3 to modulate cell fate in budding yeast. Genes Dev. 15, 2803–2808 Tyers, M. et al. (1993) Comparison of the Saccharomyces cerevisiae G1 cyclins: Cln3 may be an upstream activator of Cln1, Cln2 and other cyclins. EMBO J. 12, 1955–1968 Cross, F.R. et al. (2002) Testing a mathematical model of the yeast cell cycle. Mol. Biol. Cell 13, 52–70 Miller, M.E. and Cross, F.R. (2000) Distinct subcellular localization patterns contribute to functional specificity of the Cln2 and Cln3 cyclins of Saccharomyces cerevisiae. Mol. Cell Biol. 20, 542–555 Edgington, N.P. and Futcher, B. (2001) Relationship between the function and the location of G1 cyclins in S. cerevisiae. J. Cell Sci. 114, 4599–4611 Dirick, L. et al. (1995) Roles and regulation of Cln–Cdc28 kinases at the start of the cell cycle of Saccharomyces cerevisiae. EMBO J. 14, 4803–4813 Stuart, D. and Wittenberg, C. (1996) CLN3, not positive feedback, determines the timing of CLN2 transcription in cycling cells. Genes Dev. 9, 2780–2794 Chen, K.C. et al. (2000) Kinetic analysis of a molecular model of the budding yeast cell cycle. Mol. Biol. Cell 11, 369–391 McInerny, C.J. et al. (1997) A novel Mcm1dependent element in the SWI4, CLN3, CDC6, and CDC47 promoters activates M/G1-specific transcription. Genes Dev. 11, 1277–1288 MacKay, V.L. et al. (2001) Early cell cycle box-mediated transcription of CLN3 and SWI4 http://tig.trends.com

TRENDS in Genetics Vol.18 No.9 September 2002

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

contributes to the proper timing of the G1-to-S transition in budding yeast. Mol. Cell. Biol. 21, 4140–4148 Koch, C. et al. (1996) Switching transcription on and off during the yeast cell cycle: Cln/Cdc28 kinases activate bound transcription factor SBF (Swi4/Swi6) at Start, whereas Clb/Cdc28 kinases displace it from the promoter in G2. Genes Dev. 10, 129–141 Baetz, K. and Andrews, B. (1999) Regulation of cell cycle transcription factor Swi4 through auto-inhibition of DNA binding. Mol. Cell. Biol. 19, 6729–6741 Baroni, M.D. et al. (1994) Repression of growth-regulated G1 cyclin expression by cyclic AMP in budding yeast. Nature 371, 339–342 Tokiwa, G. et al. (1994) Inhibition of G1 cyclin activity by the Ras/cAMP pathway in yeast. Nature 371, 342–345 Flick, K. et al. (1998) Regulation of cell size by glucose is exerted via repression of the CLN1 promoter. Mol. Cell. Biol. 18, 2492–2501 Barral, Y. et al. (1995) G1 cyclin turnover and nutrient uptake are controlled by a common pathway in yeast. Genes Dev. 9, 399–409 Li, F.N. and Johnston, M. (1997) Grr1 of Saccharomyces cerevisiae is connected to the ubiquitin proteolysis machinery through Skp1: coupling glucose sensing to gene expression and the cell cycle. EMBO J. 16, 5629–5638 Carlson, C.R. et al. (1999) Regulation of the start of DNA replication in Schizosaccharomyces pombe. J. Cell Sci. 112, 939–946 Martín-Castellanos, C. et al. (2000) The puc1 cyclin regulates the G1 phase of the fission yeast cell cycle in response to cell size. Mol. Biol. Cell 11, 543–554 Forsburg, S.L. and Nurse, P. (1994) Analysis of the Schizosaccharomyces pombe cyclin puc1: evidence for a role in cell cycle exit. J. Cell Sci. 107, 601–613 Baum, B. et al. (1997) Control of S-phase periodic transcription in the fission yeast mitotic cycle. EMBO J. 16, 4676–4688 Martín-Castellanos, C. et al. (1996) B-type cyclins regulate G1 progression in fission yeast in opposition to the p25rum1 cdk inhibitor. EMBO J. 15, 839–849 Mondesert, O. et al. (1996) Cig2, a B-type cyclin, promotes the onset of S in Schizosaccharomyces pombe. Mol. Cell. Biol. 16, 1527–1533 Grallert, B. et al. (2000) A fission yeast general translation factor reveals links between protein synthesis and cell cycle controls. J. Cell Sci. 113, 1447–1458 Young, P.G. and Fantes, P.A. (1987) Schizosaccharomyces pombe mutants affected in their division response to starvation. J. Cell Sci. 88, 295–304 Kanoh, J. and Russell, P. (1998) The protein kinase Cdr2, related to Nim1/Cdr1 mitotic inducer, regulates the onset of mitosis in fission yeast. Mol. Biol. Cell 9, 3321–3334 Breeding, C.S. et al. (1998) The cdr2 + gene encodes a regulator of G2/M progression and cytokinesis in Schizosaccharomyces pombe. Mol. Biol. Cell 9, 3399–3415 Wu, L. et al. (1996) Spatial organization of the Nim1–Wee1–Cdc2 mitotic control network in Schizosaccharomyces pombe. Mol. Biol. Cell 7, 1749–1758 Suda, M. et al. (2000) Regulation of Wee1 kinase in response to protein synthesis inhibition. FEBS Lett. 486, 305–309 Kovelman, R. and Russell, P. (1996) Stockpiling of Cdc25 during a DNA replication checkpoint arrest

485

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

in Schizosaccharomyces pombe. Mol. Cell. Biol. 16, 86–93 Aligue, R. et al. (1997) Regulation of Schizosaccharomyces pombe Wee1 tyrosine kinase. J. Biol. Chem. 272, 13320–13325 Daga, R.R. and Jimenez, J. (1999) Translational control of the Cdc25 cell cycle phosphatase: a molecular mechanism coupling mitosis to cell growth. J. Cell Sci. 112, 3137–3146 Sveiczer, A. et al. (2001) A stochastic, molecular model of the fission yeast cell cycle: role of the nucleocytoplasmic ratio in cycle time regulation. Biophys. Chem. 92, 1–15 Tyson, J.J. et al. (2001) Network dynamics and cell physiology. Nat. Rev. Mol. Cell Biol. 2, 908–916 Decottignies, A. et al. (2001) In vivo localisation of fission yeast cyclin-dependent kinase cdc2p and cyclin B cdc13p during mitosis and meiosis. J. Cell Sci. 114, 2627–2640 Lopez-Girona, A. et al. (1999) Nuclear localization of Cdc25 is regulated by DNA damage and a 14-3-3 protein. Nature 397, 172–175 Rhind, N. et al. (1997) Cdc2 tyrosine phosphorylation is required for the DNA damage checkpoint in fission yeast. Genes Dev. 11, 504–511 Rupeš, I. et al. (2001) G2/M arrest caused by actin disruption is a manifestation of the cell size checkpoint in fission yeast. Mol. Biol. Cell 12, 3892–3903 Moreno, S. and Nurse, P. (1994) Regulation of progression through the G1 phase of the cell cycle by the rum1+ gene. Nature 367, 236–242 Yamaguchi, S. et al. (1997) A WD repeat protein controls the cell cycle and differentiation by negatively regulating Cdc2/B-type cyclin complexes. Mol. Biol. Cell 8, 2475–2486 Kitamura, K. et al. (1998) Fission yeast Ste9, a homolog of Hct1/Cdh1 and fizzy-related, is a novel negative regulator of cell cycle progression during G1-phase. Mol. Biol. Cell 9, 1065–1080 Lew, D.J. (2000) Cell-cycle checkpoints that ensure coordination between nuclear and cytoplasmic events in Saccharomyces cerevisiae. Curr. Opin. Genet. Dev. 10, 47–53 McMillan, J.N. et al. (1999) the morphogenesis checkpoint in Saccharomyces cerevisiae: cell cycle control of Swe1p degradation by Hsl1p and Hsl7p. Mol. Cell. Biol. 19, 6929–6939 Sreenivasan, A. and Kellogg, D. (1999) The Elm1 kinase functions in a mitotic signaling network in budding yeast. Mol. Cell. Biol. 19, 7983–7994 Russell, P. et al. (1989) Conservation of mitotic controls in fission and budding yeast. Cell 57, 295–303 Booher, R.N. et al. (1993) Properties of Saccharomyces cerevisiae wee1 and its differential regulation of p34CDC28 in response to G1 and G2 cyclins. EMBO J. 12, 3417–3426 Cross, F.R. (1988) DAF1, a mutant gene affecting size control, pheromone arrest, and cell cycle kinetics of Saccharomyces cerevisiae. Mol. Cell. Biol. 8, 4675–4684 Nash, R. et al. (1988) The WHI1+ gene of Saccharomyces cerevisiae tethers cell division to cell size and is a cyclin homolog. EMBO J. 7, 4335–4346 McAdams, H.H. and Arkin, A. (1999) It’s a noisy business! Genetic regulation at the nanomolar scale. Trends Genet. 15, 65–69 Sveiczer, A. et al. (1999) Mitotic control in the absence of cdc25 mitotic inducer in fission yeast. J. Cell Sci. 112, 1085–1092